Our long-term goal is to improve lung cancer mortality by prospectively identifying individuals at increased risk for developing lung carcinoma. This will allow opportunities for intervention by chemoprevention strategies or for the early detection of resectable tumors. The central hypothesis that forms the basis of this proposal is that novel methods for detecting and analyzing biomarkers of DNA exposure to mutagens and of DNA damage may identify those individuals with an increased susceptibility to developing DNA damage, and subsequently lung carcinoma. Our approach will use a semiquantitative LOH detection method previously described by us and a semiquantitative immunocytochemical method of DNA adduct analysis developed by Dr. Regina Santella at Columbia University to accomplish the following aims: 1) Define the relationship between biomarkers of DNA exposure to mutagens and DNA damage, as measured by DNA adduct formation and by loss of heterozygosity (LOH). We will evaluate genetic and dietary factors that modulate DNA adduct formation and LOH in a population of middle aged smokers. 2) Develop methods to assay DNA adduct formation and LOH in epithelial cells obtained noninvasively by sputum induction to allow extrapolation of these assays to larger population studies. We estimate that a total of 120 moderate smokers will be required to be enrolled at Columbia University to adequately power the study to detect an effect of family history, micronutrient intake on LOH or adduct formation. We will compare the quantity of DNA adducts of PAH and 4-ABP, extent of LOH as measured by fractional allelic loss, and the frequency of p53 mutation present in the epithelial cells of women to men.